2. Two tests are used to evaluate the level of serum potassium (mmol/L) of 10 samples, and the data are shown in Table 18.18. The Pearson correlation coefficient is r =0.925 and p<0.001.

(a) Suppose we want to assess the agreement between the two tests, is the result of correlation analysis appropriate to conclude that this is a good agreement?

Why?

(b) Suppose we want to test the difference between the readings of two tests, what test method can be used? Can we use this result to evaluate agreement?

(c) How do you determine agreement between the two tests? What is your conclusion.